Myristica*] FLORA TNDICA. 155
often not quite terminal, and the seed is then slightly oblique, the dorsum or non-
rhaphal surface being the most convex. The middle coat is hard, woody, and brittle,
and consists of a single layer of prismatic cells radiating from within outwards. The
innermost coat, which is probably the nucleus, is thin and fleshy, and gives off the
numerous irregularly branching, much anastomosing plates which divide the albumen.
These are largest near the chalaza, from which they appear to spring when the seed
is cut vertically. The albumen is composed of hexagonal cellular tissue, with thin
transparent walls, each cell enclosing an opaque mass which nearly fills its cavity.
The cells of the processes of the endosperm are smaller and darker coloured than
those of the albumen.
This is a very tropical Order, usually forming handsome trees, often with a lofty,
undivided trunk, and horizontal, more or less verticillate tranches. In India none
of the species are known further north than 26° N. lat., on jthe northern face of the
Khasia hills. From America only thirteen are described by Mr. Bentham, in a general
enumeration of the American species, so that the Order is eminently Indian. The
species are probably most numerous in the eastern part of the Malayan Archipelago.
A few are found in tropical New Holland, but none, so far as is>nown, in China.
From Africa no species have been described, hut in the British Museum there are
two specimens marked "Myristica /*" One of these, from Cape Coast, collected by
Brass, is a subscandent stipulate plant, apparently belonging to Mahacea or Euphor-
biacea, but' the other (brought by Afeelius from Sierra Leone) is in fruit, and, judg-
ing from the general aspect, probably belongs to this Order.
As Nutmegs are generally lofty trees, inhabiting dense forests, and are almost in-
variably dioecious, many df the species are very imperfectly known ; most frequently
one sex only exists in herbaria, or, if the flowers of both sexes he known, the fruit is
perhaps a desideratum. Great caution is necessary in identifying fruiting and flower-
ing specimens gathered at different times. Of many of the species we have only seen
single specimens, and have no means of determining the amount of variation to which
they are subject. We have also had few opportunities of observing this family in a
living state, but we think it probable that the shape of the leaves will be found to vary
very much, aud that it ought to be used with great caution as a specific character.
For these reasons we attach but little importance to the diagnoses and descriptions
here given. In one or two cases only have our materials been sufficient to enable us
to offei an opinion on .the limits of species ; in general we have been obliged to
content ourselves with describing as accurately as possible the individual specimen
Myiist'ica? sesquipedalis, Wall. Cat. 6809 ! is, as Dr. Wallich himself suspected,
a laurel, as is also M. glaucescens, Wall. Cat. 6790 ! M. Finlegwniana, Wall. Cat.
6793 ! is a species of Melodontm. Wall. Cat. 9017, referred doubtfully to Myristica>
.must also be excluded, as* it certainly does not belong here.
There are no doubt several very distinct genera among Nutmegs, bat the structure
(especially that of the female flowers) is so very little known, that the time has not
yet come ^establishing these genera on a secure basis. We therefore follow Blume
and Bentham in retaining the genus Myristica for the present entire, and in divid-
ing it into sections according to the modifications of the andrcecium.
Sect. 1. KNEMA. — Calyx trilobus vel tripartite, extus tomentosua.
Columna staminea apice in discum raargine antheriferum dilatata.
Stigma dilatatum, margiue pluridentatum. Cotyledons* plan®. —
ti ad apicem pedimculi axillaris a&breviati.
The species of Knema, form; on the whole^ a well-marked group, distinguishable